Traffic Signal Optimization
Omnisight’s FusionSensor is designed to give cities and communities the means to improve adaptive signal timing management. By fusing radar and video, leveraging machine learning, and real-time data processing, these sensors make traffic systems more accurate, reliable, and efficient.
Fixed Timing Problems
Traditional traffic signals operate on predetermined patterns that can't respond to real-time conditions, costing cities in efficiency, safety, and resident frustration.
Traditional traffic lights follow predetermined patterns set during off-peak hours, then run the same cycle whether there's one car or fifty waiting at the intersection. Conditions on the ground change every minute; the signal doesn't.
Drivers idle at empty intersections, watching green lights cycle on cross streets with no traffic. The cumulative effect across a corridor: longer commutes, wasted fuel, higher emissions, and growing frustration with city traffic management.
Most cities operate signals without real-time data on volumes, queue lengths, pedestrian wait times, or signal effectiveness. Engineers tune timing based on annual studies, then hope conditions don't change in between.
Crashes, special events, weather, and construction all create traffic surges that fixed timing plans can't accommodate. By the time someone notices the backup and adjusts the timing manually, the cascade is already several blocks deep.
Corridor Results
A city deployed FusionSensor across a multi-intersection arterial where commuters faced predictable bottlenecks during morning and evening peaks. HD3D radar measures vehicle counts, speeds, queue lengths, and turning movements at each intersection in real time and feeds the data continuously into the traffic management center.
Within weeks, signal timing was adjusted based on actual demand patterns instead of static schedules. Drivers stopped idling at empty intersections, commute times dropped, and resident complaints about that corridor declined sharply.
Learn More
When concerts, sporting events, weather, or roadwork create traffic surges, fixed signal plans crumble and backups cascade through the network. FusionSensor catches the surge the moment it begins and triggers signal timing adjustments at upstream and downstream intersections automatically.
The same continuous data feeds incident dashboards for the city's TMC, giving operators full visibility into what's happening and what the signal system is doing in response, instead of guessing based on driver call-ins or social media.
Signal Control Technology
FusionSensor pairs HD3D radar with AI vision and fuses both on-device through TrueEdge, then plugs straight into existing NEMA TS1 and TS2 signal controllers for adaptive timing across every approach.
Wide field of view across multiple lanes with depth, speed, and range detection. Operates day or night and through fog, snow, and other weather conditions where cameras alone fall short. 120 degree field of view, up to six lanes, all weather rated.
1080p AI vision classifies every road user the moment they approach the intersection. Vehicles, pedestrians, cyclists, motorcycles, all detected and counted in real time so signal timing can respond to the full traffic mix instead of just the cars.
Radar and vision are fused on the device 20 times per second through TrueEdge. Detection and signal adjustments run on board with no cloud round trip, no latency. Reliable real-time data, processed at the edge of the intersection.
Drop into existing intersections without rip-and-replace. FusionSensor supports SDLC for NEMA TS2 controllers (advanced stop bar detection, pedestrian presence, zone detection), NTCIP for smart-city devices, and contact closure for legacy NEMA TS1 cabinets. Adaptive timing works with the infrastructure you already have.
CAPABILITIES
FusionSensor detects pedestrians, cyclists, and large groups at every crosswalk approach. Signal timing extends automatically when groups form or vulnerable users are still in the intersection, then returns to normal flow once the crossing is clear. Crosswalks adapt to the people using them in real time.
Green time grows when demand spikes and shrinks when an approach is empty. FusionSensor measures actual volumes at every intersection 20 times per second, so cycle lengths and phase splits respond to real conditions instead of running the same plan from 7 AM to 7 PM.
Engineers no longer have to choose between optimizing for AM peak or PM peak. The signal adapts continuously, every minute of the day.
Detect queues forming at every intersection in real time. Signal timing adjusts at the corridor level to clear queues before they grow into upstream intersections or block adjacent turn lanes, keeping traffic moving smoothly through the network.
Detect approaching buses, fire trucks, and ambulances and trigger signal preemption automatically. Transit signal priority moves buses through intersections on schedule. Emergency preemption clears the path for first responders within seconds, then returns the signal to normal operation.
Sync multiple intersections along an arterial for green-wave progression based on actual traffic, not fixed timing plans. Drivers hit fewer red lights, fuel consumption drops, and corridor throughput rises without adding lanes.
Talk to our engineering team about deploying FusionSensor for adaptive signal timing on your corridor. We'll walk through your existing infrastructure, your traffic data needs, and how AST fits into your signal modernization roadmap.
Adaptive Signal Timing is a traffic control system that changes traffic light patterns based on what's happening on the road at any given moment. Unlike traditional traffic signals that follow pre-set schedules, AST continuously collects data across multiple modes of transportation to adjust lights as conditions change.
FusionSensor pairs HD3D radar with AI vision to measure vehicle counts, queue lengths, pedestrian presence, and approach speeds 20 times per second. It feeds that data into your signal controller, which adjusts timing in real time to match the live scene.
Yes. FusionSensor supports SDLC for NEMA TS2 controllers (with advanced stop bar, pedestrian presence, and zone detection), NTCIP for smart-city devices, and contact closure for legacy NEMA TS1 cabinets. No rip-and-replace required.
Yes. Adaptive signal timing reduces congestion by adjusting green lights to match real-time traffic needs. This allows for a green wave: vehicles passing through multiple green lights at intersections, reducing idling and stops.
The success of an adaptive system hinges on the quality of its detection. Sensors must reliably capture real-time data on vehicle presence, speed, and movement. Increasingly, agencies use multimodal sensors like the Omnisight FusionSensor, which provides continuous data on vehicles, bicycles, and pedestrians, especially valuable in urban corridors, university campuses, and mixed-use areas where traffic varies throughout the day.
Agencies that have implemented adaptive systems report reduced travel times, lower emissions from less idling, more consistent travel times, better accommodation of special events or seasonal patterns, and enhanced pedestrian safety through responsive crosswalk timing. In the long run, adaptive systems help cities get more out of their existing infrastructure without costly widening or reconstruction.
Detection at the intersection tells the controller a vehicle is present: historically inductive loops cut into the pavement, increasingly radar and AI vision watching the stop bar. The controller uses that presence data to decide when to serve each approach.
ASCT is the umbrella term for signal systems that adjust timing continuously from live detection instead of running fixed schedules. Signal timing responds to actual demand every cycle rather than to a plan written months earlier.
Retiming is a periodic manual update: an engineer studies the intersection and writes a new fixed plan that then ages until the next study. Adaptive control replaces that cycle by adjusting continuously from real-time data.
ATSPMs are metrics computed from high-resolution controller and detector data, such as arrivals on green and split failures, that show how well each intersection is actually serving traffic. They require reliable detection as input, which is what FusionSensor provides.
Yes, when the detection is multimodal. FusionSensor classifies pedestrians and cyclists separately from vehicles, so the controller can extend crossing windows when groups are waiting and count every road user, not just cars.